Onychomatricoma (OM) is a rare, benign, fibroepithelial neoplasm of the nail matrix predominantly affecting women with a peak incidence at around 50 years of age and presents twice as often on the fingernails than toenails.1–4 Its 4 cardinal clinical signs are nail plate thickening with leukoxanthonychia, rectilinear and parallel ridges and grooves with splinter hemorrhages, longitudinal and transverse nail plate overcurvature, and, at the frontal view of the free edge of the distal nail plate, multiple cavities.1–5 The cavities are visible by naked eye in about one-third of cases (36.7%).2 Dermoscopic evaluation of the free edge of the distal nail plate seems to offer a better visualization of this finding, formally referred to as holes, honeycomb-like cavities, or woodworm-like cavities.1,2,6,7 It has been claimed that ultrasonography, magnetic resonance imaging, and optical coherence tomography can be of great value for the diagnosis of OM.8 However we think, like others,6,7 that the histological evaluation of nail clipping offers a better visualization of the cavities of OM than such costly, when available, imaging techniques.
The definitive diagnosis of OM is made by histopathology. Macroscopic evaluation of the avulsed nail plate show specific findings as cavitations both at the proximal and the distal borders of the thickened nail plate.3 The complex 3-dimensional fibroepithelial core of OM produces 2 main histologic patterns (Fig. 9 Plates 1 and 2 of Ref. 3 and Fig. 1 Plates 1–4 of Ref. 9): (1) The lobulated or foliated pattern is observed in transverse sections with prominent deep epithelial grooves centered by V-shaped empty spaces corresponding to the negative image of the ventral spurs of the nail plate. These proximal ungual spurs covered by the keratogenous zone cell layer correspond distally to the walls dividing the cavities. (2) The pedunculated pattern is observed in longitudinal section. The fibroepithelial tumor has either a “glove-finger” monodigitate pattern or a pseudocondylomatous, multidigitate pattern. Of note, the fibroepithelial core of OM, if cut in tangential sections, will show a biphasic tumor with a fenestrated epithelial proliferation into the underlying mesenchymal component.3
Histologic evaluation of the longitudinal sections of the nail plate permits to give the precise location of the tumor.3,9,10 When the tumor involves the apical matrix and the ventral surface of the proximal nail fold (apical OM), the proximal border of the nail plate is extraordinarily thickened with multiple deep notches alternating with apical ungual spurs. Of note, the empty spaces of the notches are the negative image of the fibroepithelial digitations. OM involving the ventral matrix reveals, in longitudinal sections, a nail plate with a tapered proximal border and then a thickened nail plate with cavities (Fig. 9 Plates 1 and 2 of Ref. 3). CD99, CD10, and epithelial membrane antigen (EMA) antibodies help differentiate OM and digital tumors such as superficial acral fibromyxoma and its variants cellular digital fibroma and acquired reactive digital fibroma whose histology recalls the fibroepithelial core of OM.3 OM is characterized by a homogenous expression of CD34 and a negative or patchy expression of CD10, whereas CD99 and EMA are negative. By contrast, superficial acral fibromyxoma shows, beside CD34 expression, a homogeneous reactivity for CD99 and CD10 and focal reactivity for EMA.
The diagnosis of OM involves consideration of the 2 other onychogenic tumors presenting as localized longitudinal pachyonychia with same distal cavitation pattern of OM: onychocytic matricoma (OCM)4,10–15 and onychocytic carcinoma (OC).16,17 OCM and OC are epithelial tumors lacking a stromal component. OCM is characterized by endokeratinization with whorls of prekeratogenous and keratogenous cells. Unlike OCM, OC is characterized by the presence of keratinocytic atypia. Transverse sections of nail clipping show in these 3 tumors an honeycomb-like pattern of multiple cavities of different sizes in a thickened nail plate.4,10,17 When the classic clinical findings are present, with the corresponding nail clipping changes, the diagnosis of onychogenic tumor can be established with confidence.4,17 It has been recently suggested, as preliminary report,4,17 that the size of the cavities could be a diagnostic clue in onychogenic tumor. In OC, the cavities are uniformly less than 0.13 mm in diameter. In OM and OCM, the cavities are larger and vary in size. Cavities up to 0.15 × 0.40 mm are seen in OCM and up to 2 × 2.40 mm in OM. These histologic findings could provide a minimally invasive alternative to a nail matrix biopsy. Some authors suggested that thickening of the nail plate, splinter hemorrhages, and nail cavities can be also seen in a minority of cases of squamous cell carcinoma (18% in a series of 44 cases).5 However, these authors use a new dermoscopic term, nail pitting, for the cavities suggesting small punctate depression. By contrast, dermoscopic en face free-margin view of conventional OM is characterized by concentric spaces containing white to yellowish serous material and some black dots. Furthermore, nail clipping was not analyzed in this study. Thus, the author believe that these cases of Teysseire et al should be interpreted as Bowen disease of the nail bed with longitudinal subungual keratosis or OC, a rare reported and perhaps underrecognized entity.
Recently, the author has been impressed by a minute papillomatous appearance of a new variant of OM, which poses, on nail clipping, a challenging differential diagnosis with OCM. In this study, we documented 4 cases, descriptively named onychomatricoma micropapilliferum (vs. the conventional OM macropapilliferum) to highlight their characteristic clinicopathologic findings.
REPORT OF CASES
A 55-year-old white man presented with a 5-year history of lesion involving the entire nail plate of the fifth left toe. The nail plate was yellowish, thickened, and transversely overcurvated with ridging and swelling of the proximal nail fold. This suggested either OM or verrucous squamous cell carcinoma. The lesion was surgically removed with an en bloc excision of the nail unit including the 2 lateral nail folds, the proximal nail fold and the hyponychium. The nail unit was cut transversely and divided into 3 zones including the root of the nail, the lunular region, and the proximal nail bed. The hyponychium area was cut in longitudinal sections.
Macroscopic evaluation of the nail unit did not found the conventional cavitation pattern of OM (Figs. 1A, B). Frontal view of the nail showed a free edge wall-like pattern with pitting (Fig. 1A). At the level of the lunular region (Fig. 1B), the yellow nail plate was homogenously thickened. Its ventral margin was underlined by a thin white epithelium with a serrated pattern. Some white dots were visible in 1 side of the nail plate. Histologically, the most proximal sections showed a fibrillary stroma surrounding both ventrally and dorsally the proximal epithelial extension of the apex of the matrix (Fig. 2A). This proximal epithelial extension, recently called the epithelial tail of the apical matrix,18 was papillomatous. Under the proximal nail fold, a papillated epithelium with keratogenous zone cells layer covers both the proximal portion of the ventral portion of the proximal nail fold (eponychium) and the ventral matrix. Numerous superficial epithelial upward extensions were seen into the thin nail plate without cavitation. In the lunular region (Fig. 2B), the stroma of the OM replaces the entire matrical nail dermis without involvement of the dermis of the lateral nail fold. The thickened nail plate presented either elongated thin projection of dermal papillae covered by a thick matrical epithelium or tiny nests of eosinophilic onychocytes. The stroma of OM extended into the proximal nail bed. At this level, the papillomatosis was prominent with broad papillae. The distal nail bed was free of stroma and showed epithelium hyperplasia with a discontinuous granular layer.
The nail plate thickness and average dimension of the cavities were measured at the distal edge of the nail plate with a micrometer. The distal nail thickness was 5 mm. Owing to the longitudinal sections, only the vertical axis of the cavities was available with an average dimension of 0.16 mm (between 0.05 and 0.40). Near the free edge of the nail plate, the subsidence of the walls of the cavities formed open cavities with large pools of serum limited by nail scales (Fig. 2C). These modified ruined cavities appeared clinically as free edge nail pitting. The stroma was evaluated with antibodies against CD34 (clone Qbend 10, prediluted; Dako, Trappes, France), CD10 (clone 56C6prediluted; Dako), S-100 protein (polyclonal, prediluted; Dako), CD99 (clone12E7, dilution 1/50; Dako), and EMA (clone E29, prediluted; Dako) using the Envision System Flex+ (Dako). CD34 was diffusely and strongly expressed. CD10 expression was restricted to rare perivascular fibroblasts. EMA, S-100 protein, and CD99 were negative.
A 33-year-old black woman presented with a 3-year history of an asymptomatic pigmented band of the fifth right toe (Fig. 3A). A physical examination revealed a total melanonychia from the proximal nail fold to the free margin of the nail plate (consistent with ethnic nonmelanoma Hutchinson sign), associated with a thickened yellowish nail plate and increased transverse and longitudinal overcurvature. The nail plate surface showed white dots, multiple irregular longitudinal white striations, and nail scales. Frontal view showed an evenly thickened nail plate without woodworm-like holes (Fig. 3A). Dermoscopy of the distal overcurvated nail plate showed multiple irregular white lines alternating with longitudinal grooves and white dots, whereas the distal free edge was thickened without holes (free edge nail wall-like pattern without pitting) (Fig. 3B). Nail clippings were obtained with a double-action nail splitter. Hematoxylin–eosin (H&E) stain of nail clipping showed a pigmented nail plate with multiple small cavities. Such cavities were filled with serous fluid or rarely hematic brown deposits (Fig. 4A). Some of these cavities were surrounded by onion-like lamellation of keratogenous cells (Fig. 4B), reminiscent of the retarded maturation pattern observed in the nail plate of the keratogenous subtype of OCM.11 The melanin pigment was evenly distributed without melanocytes. Periodic acid–Schiff (PAS) staining was negative for fungi and highlighted serous contents in each cavity (Fig. 4C). The nail plate was 2.40 mm thick. Measuring between 0.16 × 0.08 and 0.20 × 0.12 mm (average of 0.13 mm on the vertical axis and 0.10 mm on the transverse axis), the cavities suggested a pigmented OCM. However, excision of the lesion was asked to rule out an OC.
The 2 specimens received, the avulsed nail plate and the excised tumor, were cut transversely. The avulsed nail plate was divided within 4 zones including the root of the nail plate, the lunular region, the nail bed, and the distal edge. In the most proximal sections of the avulsed nail plate (Fig. 5A), epithelial papillomatosis started at the most proximal part of the matrix sparing the proximal nail fold. Owing to antero-oblique orientation, the superficial epithelial elevations were cut numerous times in cross-section during their tortuous course through the thin nail plate. The papillated epithelium formed spheres of clear prekeratogenous zone cells with vesicular nuclei and eosinophilic keratogenous cells layers with pycnotic hyperchromatic nuclei. The spheres centered minute papillae or empty spaces. At the lunula, there are numerous rings of eosinophilic keratogenous cell layers centering either clear onychocytes or nests of eosinophilic onychocytes intermingled with shadow cells in the thickened nail plate (Fig. 5B). Several eosinophilic spheres contain a central tiny cavity with coagulated plasma and dispersed mononuclear cells.
The distal nail bed showed a discrete epidermal papillomatosis with a discontinuous granular layer. The nail plate was thickened (Fig. 5C) with numerous tiny cavities. Along the nail plate, superficial fissures explain the irregular grooves clinically observed. Some microcavities, best visualized with PAS stain (Fig. 4C), approach the surface of the nail plate and correspond to the white dots observed clinically.
In the fibroepithelial tumor core (Fig. 6A), a papillomatous epithelium with some elongated dermal papillae form minute V-shaped empty spaces corresponding to the negative image of the papillomatous prekeratogenous and keratogenous zones found attached to the nail plate. Proximally, the typical stroma of OM was organized in 2 layers infiltrated by mast cells. In the superficial layer, thin wavy fibroblast nuclei were randomly oriented, whereas the deep layer showed fibroblasts and thicker collagen bundles mainly oriented along a horizontal axis. Distally, the connective tissue core formed a large nodule with spindle fibroblasts arranged in a random and loose storiform pattern. CD34 was diffusely and strongly expressed in the stroma (Fig. 6B), whereas CD10 was focally and predominantly expressed in perivascular fibroblasts. EMA, S-100 protein, and CD99 were negative.
A 60-year-old white man presented a tumor of the right great toe located of the lateral part of the nail plate. This suggested either a cutaneous horn or a Bowen disease as longitudinal subungual keratosis. An excisional lateral longitudinal biopsy was performed.
The sample was cut longitudinally. It showed the classic stroma of OM involving the entire proximal nail fold including the angle of reflection between the ventral and dorsal portion of the proximal nail fold (Fig. 7). The eponychium showed a matrix metaplasia with exophytic epithelial proliferation. At the inferior portion of the eponychium, the broad or elongated papillae were separated from the nail plate by cut artifact and formed minute V-shaped empty spaces (Fig. 7A). The papillated prekeratogenous and keratogenous zone remained adherent to the proximal border of the nail plate. This papillomatous proximal border was extraordinarily thickened without the deep notches alternating with apical ungual spurs of the conventional apical type of OM. Of note, the nail plate presented multiple cleft artifacts as longitudinal pseudoholes that differed from the tunnel of the classical type of OM by the fraying of the nail plate within the pseudocavities (Fig. 7A). The nail plate thickness was 4.45 mm. The cavities were small, rare, and measured between 0.03 and 0.15 mm in their vertical axis: (average of 0.10 mm). Proximally, the voluted nail plate showed multiple superficial epithelial upward extensions (Fig. 8A), immediately followed by oval nests of eosinophilic onychocytes and shadow (ghost) cells (Fig. 8B). More distally, these onychocytes had a clear cytoplasm with distinct borders (Fig. 8C). Keratohyaline granules were focally seen in the ventral matrix. The nail bed epithelium manufactured focally a thin epidermal cornified layer. There was a diffuse and strong expression of CD34 by the stroma, whereas only rare perivascular fibroblasts expressed CD10. EMA, S-100 protein, and CD99 stains were negative.
A 66-year-old white man presented with an asymptomatic pigmented band on the right great toenail. On physical examination, a 5-mm longitudinal melanonychia was associated with nail dystrophy and proximal nail fold edema. Radiography revealed no bone involvement. Clinically, 3 tumors were considered: melanoma, squamous cell carcinoma, and keratoacanthoma.
Surgical excision provided 2 specimens, the avulsed nail plate and the excised tumor. Macroscopic evaluation of the nail plate did not reveal the ordinary proximal and distal cavitation pattern of OM. The nail plate was homogenously thickened. Its proximal margin was undulated and underlined by a thin white epithelium. The 2 specimens were cut longitudinally. Histologic spatial reconstitution demonstrated an OM involving the ventral surface of the proximal nail fold because the proximal border of the nail plate was extraordinarily thickened (Fig. 9A). The papillomatous prekeratogenous and keratogenous zones, which adhered to this proximal edge, lacked the deep notches alternating with apical ungual spurs of the conventional apical OM. The nail plate thickness was 2.65 mm. The cavities were small, rare, and measured between 0.10 and 0.15 mm in their vertical axis: (average of 0.12 mm). The ventral matrix showed several zones of epidermal metaplasia. The excised tumor had the classical stroma of OM. The eponychium had an acanthotic and papillomatous epithelium with a thick prekeratogenous zone cells where eosinophilic spheres of keratogenous cells were centered by clear onychocytes (Fig. 9B). In some tangential sections, the spheres were almost made of onion-like layers of eosinophilic onychocytes (Fig. 9C). These eosinophilic spheres were surrounded by basophilic germinative cells with peripheral palisading and areas of retraction of the stroma from tumor islands resulted in peritumoral lacunae. Therefore, these spheres mimicked horn cysts with hyaline keratin masses of keratotic basal cell carcinoma. However, close examination of these pseudohorn cysts showed onychocytic shadow cells. A keratotic basal cell carcinoma with pilomatricomal differentiation was ruled out by the absence of BerEp4 staining (prediluted, Dako). The stroma demonstrated a diffuse and strong expression of CD34, whereas CD10 expression was restricted to rare perivascular fibroblasts. EMA, S-100 protein, and CD99 were negative.
We report 4 cases of a previously undescribed type of OM involving the ventral matrix including its apex, and the proximal nail bed in cases 1 and 2, or the apical matrix and the ventral portion of the proximal nail fold in cases 3 and 4. The clinical presentation was unusual as a thickened free edge of the nail plate without discernible cavities.
OM micropapilleferum differs from the ordinary OM by 3 features:
- The lack of cavitation at the proximal border of the nail plate and the small sizes of the cavities at the distal free edge.
- A papillated epithelial hyperplasia pattern very different from the digitate pattern of the ordinary OM.
- A matrical keratinization with pseudohorn cyst-like areas closely mirroring those found in OCM.
Furthermore, the sex ratio and sites of the lesion differ from those of conventional OM with the caveat that this series is small.
A practical approach to the diagnosis of onychogenic tumor mainly involves consideration of tumors that clinically present as localized longitudinal pachyonychia including melanoma and subungual bowen disease. Whether pachyonychia is caused by a thickened nail plate or by a localized band of subungual hyperkeratosis may not be clinically and dermoscopically obvious, and leucoxanthonychia or longitudinal melanonychia is observed in OCM, OM, and OC. Therefore, the definitive diagnosis of these 3 onychogenic tumors is made by histopathology on nail clipping specimen or nail biopsy. OM is easily diagnosed as a fibroepithelial tumor, keeping in mind its micropapilliferum variant that can simulate trichoblastoma or basal cell carcinoma on biopsies without nail plate. Indeed, when the avulsed nail plate is not received by the dermatopathologist, an unfortunately frequent practice in our consultation file, it is sometimes difficult to differentiate the pseudohorn cyst of OM micropapilleferum from horn cysts with compact hyaline keratin masses of trichoblastoma including trichoepithelioma, keratotic basal cell carcinoma, or squamous eddies of seborrheic keratosis. As previously indicated in the seminal report of OCM4,11 and perfectly demonstrated in this series, the pseudohorn cysts of both OCM and OM micropapilleferum have 2 distinct layers with a ring pattern: (1) the clear prekeratogenous zone cells with oval or round nuclei, with evenly dispersed chromatin, and (2) the keratogenous zone composed of tightly packed polygonal to fusiform eosinophilic cells with pycnotic nuclei. In addition, the transitional eosinophilic onychocytes become progressively clear with shadow cells. By contrast, horn cysts with hyaline and trichilemmal-type keratinization have rounded or irregular shapes, a thin inner layer of eosinophilic cells with large, oval, pale, vesicular nuclei, and are filled with compact keratinous masses without transition to onychocytic shadow cells. The squamous eddies of irritated seborrheic keratosis are easily differentiated from the pseudohorn cyst of OM by their inner layer of eosinophilic flattened squamous cells with oval to fusiform hyperchromatic nuclei, and their loose or compact eosinophilic keratinous masses without transition to onychocytic shadow cells.
To avoid confusion with the pseudohorn cysts of seborrheic keratosis which present a thin layer of flattened granular cells and laminated cornified cells, we propose to designate the pseudohorn cysts in OM and OCM as keratogenous spheres. The papillae of the latter end as a tip without keratogenous zone explaining the microcavities. The free edge nail wall-like pattern with pitting is explained by cavities with an average dimension of 0.16 mm (case 1), and the wall-like pattern without pitting is explained by cavities with an average dimension of 0.13 x 0.10 mm (case 2–4). The microcavities getting in touch with the surface of the nail plate are responsible for the white dots (the so-called milia cysts) observed by dermoscopy both in OCM13,14 and OM micropapilliferum. The low, projecting ridges separated by the irregular longitudinal furrows explain the clinical irregular white line. The evenly thickened free edge of the distal nail plate is either explained by the small size of the cavities or by the keratogenous zone at the tip of the papillae which manufacture a homogeneous thick nail plate. This free edge nail wall-like pattern (with or without pitted wall) is in stark contrast to the usual honeycomb-like cavities seen in conventional OM. It is inferred that these dermatoscopic findings could be clinical clues to differentiate both OCM and OM micropapilliferum from conventional OM. In the initial description of OCM, this entity was clearly differentiated from seborrheic keratosis.11 From time to time, these 2 lesions continue to pose problems in the histological differential diagnosis and OCM with its various clinical presentations as leucoxanthonychia or melanonychia has been described using different names as subungual seborrheic keratosis, nail unit acanthoma, or longitudinal subungual acanthoma.19,20 These new superfluous synonymies add confusion in nail tumors. In the estimation of the author, these so-called new entities are OCM, if the histologic criteria of keratogenous spheres defined in this article are used.
In sum, there are 2 clinicopathological variants of OM: macropapilliferum and micropapillerum. As OM micropapillerum has small cavities, the main differential diagnosis on nail clipping is OC.
This work is dedicated to the late Pr Robert Loubiere, an outstanding scientist and a special human being. Ce travail est dédié au Dr Robert Loubiere, mon Maître en médecine, mon Maître en vertu d'humanité.
1. Lesort C, Debarbieux S, Duru G, et al. Dermatoscopic features of onychomatricoma
: a study of 34 cases. Dermatology. 2015;231:177–183.
2. Di-Chiacco N, Tavares GT, Tosti A, et al. Onychomatricoma
: epidemiological and clinical findings in a large series of 30 cases. Br J Dermatol. 2015;173:1305–1307.
3. Perrin C, Baran R, Balaguer T, et al. Onychomatricoma
: new clinical and histological features. A review of 19 tumors. Am J Dermatopathol. 2010;32:1–8.
4. Perrin C, Cannata GE, Langbein L, et al. Acquired localized longitudinal pachyonychia
and onychomatrical tumors: a comparative study to onychomatricomas (five cases) and onychocytic matricomas (four cases). Am J Dermatopathol. 2016;38:664–671.
5. Teysseire S, Dalle S, Duru G, et al. Dermatoscopic features of subungual squamous cell carcinoma: a study of 44 cases. Dermatology (Basel). 2017;233:184–191.
6. Miteva M, Cadore de Farias D, Zaiac M, et al. Nail
clipping diagnosis of onychomatricoma
. Arch Dermatol. 2011;147:1117–1118.
7. Okon LG, Saedi N, Schwartz L, et al. A case of onychomatricoma
: classic clinical, dermoscopic, and nail
-clipping histologic findings. J Am Acad Dermatol. 2017;76:S19–S21.
8. Cinotti E, Veronesi G, Labeille B, et al. Imaging technique for the diagnosis of onychomatricoma
. J Eur Acad Dermatol Venereol. 2018;32:1874–1878.
9. Perrin C, Langbein L, Schweizer J, et al. Onychomatricoma
in the light of the microanatomy of the normal nail
unit: new clinical and histological features. Am J Dermatopathol. 2011;33:131–139.
10. Perrin C. Tumors of the nail
unit. A review. Part II: acquired localized longitudinal pachyonychia
and masqued nail
tumors. Am J Dermatopathol. 2013;35:693–712.
11. Perrin C, Cannata GE, Bossard C, et al. Onychocytic matricoma
presenting as pachymelanonychia longitudinal. A new entity (report of 5 cases). Am J Dermatopathol. 2012;34:54–59.
12. Spaccarelli N, Wanat KA, Miller CJ, et al. Hypopigmented onychocytic matricoma
as a clinical mimic of onychomatricoma
: clinical, intraoperative and histopathologic correlations. J Cut Pathol. 2013;40:591–594.
13. Song H, Qu F, Dang N, et al. Onychocytic matricoma
: report of an asian case. Ann Dermatol. 2017;29:355–357.
14. Kusutani N, Kamo R, Sowa-Osako J, et al. Onychocytic matricoma
as an under recognized benign mimicker of subungual malignant melanoma and Bowen's disease. J Dermatol. 2017;44:e73–e74.
15. Perrin C. Germinotropic onychocytic matricoma
: a new histopathologic subtype of onychocytic matricoma
in the light of the microanatomy of the normal nail
unit, with special reference to nail
mesenchymeExpression of follicular sheath keratins in the normal nail
with special reference to the morphological analysis of the distal nail
unit. Am J Dermatopathol. 2017;39:e97–e101.
16. Perrin C, Langbein L, Ambrosetti D, et al. Onychocytic carcinoma
: a new entity. Am J Dermatopathol. 2013;35:679–684.
17. Perrin C, Cannata GE, Ambrosetti D, et al. Acquired localized (monodactylous) longitudinal pachyonychia
and onychocytic carcinoma
in situ (two cases) Part II. Am J Dermatopathol. 2017;39:40–44.
18. Perrin C. Nail
psoriasis and nail
extensor enthesitis theory, what is the link. Am J Dermatopathology. 2018. In press.
19. Stinco G, Errichetti E, Patrone P. Ungual seborrheic keratosis : report of a case and its dermoscopic features. J Eur Acad Dermatol Venereol. 2016;30:481–483.
20. Baran R, Moulonguet I, Goetmann-Bonvallot S, et al. Longitudinal subungual acanthoma: one denomination for various clinical presentations. J Eur Acad Dermatol Venereol. 2018;32:1608–1613.